Chapter 4: Engineering transcription factor EB (TFEB) to modulate
4.3 Transcription factor EB (TFEB) mutation strategies
We have selected activation domain (AD) and DNA binding domain (DBD) of TFEB as mutation candidates to determine whether we could optimize an M1 or M2 state of macrophage, or generate a new state.
Figure 2. TFEB Mutation Experimental Design and Strategy (A) An experimental concept for rationally designed TFEB mutants and libraries at activation domain (AD) and DNA binding domain (DB) (B) Design strategy for rationally selected mutants using Position Specific Scoring Matrix (PSSM). Arrows indicate what the current amino acids at those positions are and point to the amino acids which will be substituted for those positions. High positive normalized log- likehood indicates that the amino acid is more likely at the given position.
The rationale for choosing those domains is their functional importance, in particular AD is required for its transcriptional activation function and DNA binding domain is critical to have TFEB bind to DNA in the form of homodimers and heterodimers (Napolitano, et al.
2016, Hemesath et al., 1994, Pogenberg et al., 2012). Site-saturated mutagenesis libraries and rationally-chosen mutants were generated in AD and DB domain of TFEB. In order to rationally choose mutants in AD and DB domains, we fed the TFEB sequence into Protein Structure Prediction (ProSPr) to compute normalized log-like likehood of substitution at a given position (Billings, et al. 2019). The Position Specific Scoring Matrix (PSSM) heat map is shown in Figure 2.b, where positive blue colors indicate that the amino acid’s presence is more likely to be in that position, whereas red colors indicate that amino acid’s presence is unlikely to be at that position. Three mutants were created in DNA binding domain and two in Activation Domain. Arginine in positions 245-248 was changed to the next most likely predicted amino acid, Lysine. Another mutant that would likely have a disabling effect on the DNA binding domain was generated by changing Arginine in 245- 248 to one of the least likely predicted amino acids, Cysteine. We have also chosen a mutant that creates a single-point mutation at position 258 from Leucine to Methionine (Figure 2.b). For AD, at positions 158 and 159 Aspartic Acid was mutated to the next most likely amino acid, Glutamic Acid. Similarly, at positions 168 and 169 Aspartic Acid was mutated to Asparagine. In order to make RNA libraries to transfect into cultured macrophages, TFEB was cloned into an RNA backbone containing T7 promoter, 3’untranslated region (UTR), and a poly-A tail via overlap polymerase chain reaction (PCR). PCR fragments were then made into RNA using in vitro transcription, 5’capping, and poly-A tailing kits. In vitro transcription kit with modified nucleosides was used for production of RNA to ensure macrophages can be transfected and minimize their immune response to the delivered cargo. The RNA was capped post transcription and poly-A tail extended to ensure stability.
4.4 Exploring DNA transfection for introduction of TFEB libraries
In order to generate classical M1 and M2 states in cultured human macrophages and explore whether DNA transfection is a potential route for transfecting TFEB mutants, we
Figure 3. Experimental results of DNA transfection and M1/M2 state induction in cultured human macrophages (A) Experimental design in which M1/M2 induction and plasmid transfection conditions are multiplexed into a single scRNA-seq experiment (B) Macrophage gene expression data projected into UMAP (Scanpy), colors indicate the different conditions and their source of induction (C) Top panel shows expression of markers pertaining to M2 macrophage state, whereas bottom panel shows expression of markers pertaining to M1 macrophage state.
performed a small pilot experiment as shown in Figure 3.a. One of the reasons DNA transfection is appealing is that the transfected plasmids can stick around, whereas RNA libraries can degrade, and later captured via feature capture probe on beads containing the complementary sequence (Supplementary Figure 5). We induced classical M2 state using IL4, IL10, and TGF-𝛽𝛽 (20 ng/mL) and M1 state using LPS (50 ng/mL), IFN-ᵞ (20 ng/mL).
An IL-10 (20 ng/mL) condition was included given prior literature suggesting that IL-10 can inhibit starvation induced autophagy in macrophages (Hun-Jung Park, et al. 2011).
Three plasmids were transfected in macrophages using PromoFectin transfection reagent:
TFEB with a feature capture probe (cloned between TFEB stop codon and SV40 poly-A signal), TFEB fused with GFP (TFEB-GFP) (pEGFP-N1-TFEB, addgene #38119), and mNeonGreen (pAAV-CAG-mNeongreen, addgene #99134). Plasmid structures for these are shown in Figure 4.b. TFEB-GFP was used to visualize transfection and presence of TFEB in macrophages, while mNeonGreen was used as a control for transfection. Two control wells of macrophages with no stimuli were also processed to serve as a baseline for M0 macrophage state. Upon transfecting the plasmids, no fluorescence was observed in either TFEB-GFP or mNeonGreen in macrophages. However, we still processed these conditions for scRNA-seq. All conditions were multiplexed into a single 10X chromium lane using lipid-tagged indices from MULTI-seq protocol (McGinnis, et al. 2019). In order to attempt capturing the TFEB containing the feature capture probe, we created a primer to amplify the user introduced TFEB. Capture of transfected TFEB is important for determining which TFEB libraries got expressed since macrophages produce TFEB transcripts natively. scRNA-seq data was aligned as outlined in the Methods section.
Samples were demultiplexed using MUTLI-seq indices and Scanpy was used to normalize, analyze, and visualize the data. The control samples demonstrate an overlap of cells as expected, suggesting a lack of variability in states between unperturbed macrophages (Figure 3.b). We first verified that we observed expected M1 and M2 markers in the respective conditions. TGM2, CD209, and IL17RB were all prominently expressed in M2 cluster (Martinez, et al. 2013, Buchacher, et al. 2015, Gratchev, et al. 2021). WARS, CXCL9, and GBP1 had prominent expression in M1 condition cluster (Brown, Wallet, et al. 2010; Orecchioni, et al. 2019; Qiu, et al. 2018). This confirmed delineation between the
conditions is useful in serving as a ruler on which we can project the cell states of the mutants proposed in Section 4.3 (Figure 3). Next, we used Scanpy’s function (rank gene groups via Wilcoxon Rank-Sum) to determine genes most prominent in each condition of Figure 4.a. Macrophages are notoriously challenging to transfect with DNA and we did not observe any transfection via fluorescence, however we were curious to see if there were any interesting responses induced by the different plasmids shown in Figure 4.b.
mNeonGreen is seen producing prominent antiviral proteins; MX1 as pictured in Figure 4.b and MX2 as seen in Figure 4.a (Braun, et al. 2015), and interferon-induced genes IFIT1, IFIT2 whose expression is enhanced following viral infection (Pidugu, et al. 2019). These genes are also observed in the M1 cluster, potentially hinting at the viral element present on the mNeonGreen as being responsible for inducing a pro-inflammatory response. In particular, mNeonGreen plasmid contains a CAG promoter which is composed of cytomegalovirus (CMV) early enhancer, chicken beta-actin gene, and a splice acceptor of the rabbit beta-globin gene (Miyazaki, et al. 1989, Niwa, et al. 1991). Additionally, it has inverted terminal repeats (ITR) from Adeno-Associated Viruses and woodchuck hepatitis virus posttranscriptional regulatory element (WPRE). Given that no expression was seen from mNeonGreen, it is highly likely that one or a combination of these elements contributed to macrophages starting to move into pro-inflammatory anti-viral state.
Macrophages transfected with TFEB-FC display an ambiguous state between pro- and anti- inflammatory. Macrophages in this condition expressed IL6, which can act as both a pro- inflammatory and anti-inflammatory cytokine, and IL10, which is known as an anti- inflammatory cytokine. Additionally, they expressed CXCL5, which is known recruit neutrophils and regulate inflammatory response. Interestingly, TFEB-GFP appears to highly express genes that are also observed in M1 macrophages and cells infected with mNeonGreen. Our hypothesis as to why TFEB plasmid did not induce a similar effect is due to presence of GFP in the TFEB-GFP plasmid. TFEB is natively expressed in macrophages, while GFP is foreign, making it more likely to raise an alarm. Given that the purpose of TFEB is to induce autophagy, we have screened four known autophagy markers, ULK1, ATG7, LAMP1, and DRAM1, and did not see a distinct expression of these genes in TFEB conditions (Bednarczyk, et al. 2018) as seen in Supplementary Figure 6. Overall,
the experiment confirmed challenges of macrophage DNA transfection, yielded interesting data on macrophage response to different plasmids, and provided clean M1, M2
transcriptomic data.
Figure 4. Prevalent genes expressed in macrophages transfected with TFEB and mNeonGreen DNA plasmids (A) Most prominent genes ranked for each condition using Scanpy’s rank gene algorithm (B) Transfected vectors and some of the most prominent genes expressed in those conditions visualized on UMAP
4.5 Future Work
Given a lack of success with DNA transfection in macrophages, we have generated the RNA TFEB mutant libraries described in Figure 2. We have confirmed that RNA transfection yields good results (> 80% fluorescence) in Supplementary Figure 7. The next steps are to analyze the impact of rationally designed mutants, and site-saturation mutagenesis libraries, and understand how the method of transcription factor engineering via scRNA-seq can be improved.
4.6 Acknowledgements
David Brown provided a script to de-multiplex MULTI-seq tags. Jong Hwee Park helped with the MULTI-seq portion of the protocol and provided feedback on experimental design. Elowitz lab provided RNA backbone and Citrine RNA for our development and testing.
4.7 Methods
Transcript Alignment
For transcriptome read alignment and gene expression quantification, we used 10x Cell Ranger v5.0.1 with default options to process the FASTQ files from the transcriptome sequencing library. The reads were aligned against the human reference provided by Cell Ranger (hg19-1.2.0).
Macrophage Culturing
Primary human monocytes were purchased from Stem Cell Technologies and cultured in RPMI 1640 supplemented with 10 vol % FBS with human M-CSF (Miltenyi Biotec, #130-
096-491) to induce macrophage state (50 ng/mL). Cell medium was changed every third day and the cells were washed with pre-warmed medium to remove non-adherent cells. Cells were transfected with DNA on the seventh day post-seeding and culturing with M-CSF.
DNA was transfected using PromoFectin-Macrophage (PromoCell, # PK-CT-2000-MAC- 10).
RNA Production
Site-saturation mutagenesis primer libraries were ordered from Twist Biosciences.
Individually selected primer mutants were ordered from IDT. Both libraries and individual mutants were created using overlap PCR and then 1 ug of DNA was fed into Cell Script’s INCOGNITO T7 Ψ-RNA In Vitro Transcription kit (#C-ICTY110510). The 5’ capping and poly-A extension was done using Cell Script’s T7 mScript Standard mRNA Production System kit (#C-MSC11610). All intermediate reactions were purified using MEGAclear transcription clean-up kit (ThermoFisher, #AM1908).
4.8 Supplementary Material
Figure 52. Conceptual drawing demonstrating how TFEB mutants introduced into macrophages would be captured using 10X Genomics’ Capture Sequence 2 on the Gel Bead along with the rest of the cell transcriptome
Figure 6. UMAP projection showing expression of known autophagy markers across the various conditions outlined in Figure 3.
Figure 3. Images of cultured macrophages (day 7 post M-CSF rest) showing transfection with Citrine RNA (bottom panel).
Top shows morphology of the macrophages at day 7 and post RNA transfection in bright field view.
4.9 References
Rőszer, T. Understanding the Mysterious M2 Macrophage through Activation Markers and Effector Mechanisms. Mediators of Inflammation 2015, 1–16 (2015).
Ardura, J. A. et al. Targeting Macrophages: Friends or Foes in Disease? Front. Pharmacol. 10, (2019).
Klichinsky, M. et al. Human chimeric antigen receptor macrophages for cancer immunotherapy.
Nat Biotechnol 38, 947–953 (2020).
Zhang, F. et al. Genetic programming of macrophages to perform anti-tumor functions using targeted mRNA nanocarriers. Nat Commun 10, (2019).
Puertollano, R., Ferguson, S. M., Brugarolas, J. & Ballabio, A. The complex relationship between TFEB transcription factor phosphorylation and subcellular localization. EMBO J 37, (2018).
Fang, L. et al. Transcriptional factor EB regulates macrophage polarization in the tumor microenvironment. OncoImmunology 6, e1312042 (2017).
Pastore, N. et al. TFEB and TFE3 cooperate in the regulation of the innate immune response in activated macrophages. Autophagy 12, 1240–1258 (2016).
Napolitano, G. & Ballabio, A. TFEB at a glance. Journal of Cell Science (2016) doi:10.1242/jcs.146365.
Hemesath, T. J. et al. microphthalmia, a critical factor in melanocyte development, defines a discrete transcription factor family. Genes & Development 8, 2770–2780 (1994).
Pogenberg, V. et al. Restricted leucine zipper dimerization and specificity of DNA recognition of the melanocyte master regulator MITF. Genes & Development 26, 2647–2658 (2012).
Billings, W. M., Hedelius, B., Millecam, T., Wingate, D. & Corte, D. D. ProSPr: Democratized Implementation of Alphafold Protein Distance Prediction Network. (2019)
doi:10.1101/830273.
Park, H.-J. et al. IL-10 inhibits the starvation induced autophagy in macrophages via class I phosphatidylinositol 3-kinase (PI3K) pathway. Molecular Immunology 48, 720–727 (2011).
McGinnis, C. S. et al. MULTI-seq: sample multiplexing for single-cell RNA sequencing using lipid-tagged indices. Nat Methods 16, 619–626 (2019).
Martinez, F. O. et al. Genetic programs expressed in resting and IL-4 alternatively activated mouse and human macrophages: similarities and differences. Blood 121, e57–e69 (2013).